KR20010051142A - Arrangement for resource and work-item selection - Google Patents

Abstract

PURPOSE: An arrangement method for selecting resource and work-item are provided to match the work item with the resource by supplying a peak value to all the stake holders of a workflow. CONSTITUTION: A selection engine determines available resources that possess skills needed by an available work item. Each determined resource determines both a business value(RSBV) of having that resource service the work item and a value(RTV) to that resource of servicing the work item, and then selects the resource that has a best combined value of the business value and the value to the resource to serve the work item. The business value is a scaled sum of products of a proficiency level of the resource in each of the skills and the weight of that skill of the work item. The value to a resource or work item is a scaled sum of products of scaled resource treatments or work item treatments and weights given to those respective treatments by the work item.

Description

Arrangement for resource and work-item selection}

Technical field

The present invention relates to workflow management.

Background of the invention

In a workflow, a work item--a physical item, such as a product or paper document, or a virtual item, such as a communications and electronic document--is a resource, for example a worker, wholly or partially allocated to that stage. , One or more processing steps executed by the machine and the agent, respectively. Representative workflows include call centers and assembly lines.

Traditional call-center resource and work-item selection algorithms are limited by the call-center model given by the automatic call distribution (ACD) switch. This conventional workflow generally uses a first-in, first-out (FIFO) selection algorithm in each state: the first available resource processes the first available work item. A variation on this basic algorithm has been developed over time, but the underlying algorithm remains the same. For example, one known ACD switch represents a resource as having a skill and skill level, and provides a multi-priority queue that queues work items to wait for resources in response to available technology. do. Thus, the algorithm applied to this model is limited by a limited and fixed structure. This conventional approach addresses simply deviations from the underlying FIFO selection principle. However, it quickly becomes bulky when used to implement complex or elastic selection algorithms, such as when it is necessary to select a "best" work item for a "best" resource. The limitations of this conventional approach are to sell three types of insurance, each agent must be licensed to sell each type of insurance on a state-to-state basis, and the call center allows the caller to choose an English or Spanish agent. This is evident when considering the example of an insurance company. This results in a combination of 300 possible technical requirements. Managing both conventional ACD switches is not impossible but difficult for all of these combinations.

The present invention is presented to solve these and other problems and inconveniences of the prior art. In general, in accordance with the present invention, a new workflow resource-selection and work item-selection arrangement is provided that attempts to match a work item with a resource in such a way as to bring a maximum value to all stakeholders in the workflow. . This includes customers, resources, managers, and so on. The arrangement defines the resources as having conventional skills, preferably also having a level of proficiency in these technologies, as in the prior art. It also classifies the work item as requiring a particular skill when arriving for processing, as is conventional. Under resource surplus conditions, the arrangement determines the available resources that own the technology required by the available work items, and for each resource determined, the business value and service value to the work item that causes the resource to service the work item. Determine the value of the resource that provides the resource, and then select the resource with the best combination of business value and the value as the resource that serves the work item. Under work item surplus conditions, the arrangement determines the available work items that require skills owned by the available resources, and for each determined work item, the business values and resources that enable the work item to be serviced by the resource. Determine the value of the work item for which the service is provided by, and then select the work item with the best combination of business value and said value as the work item for which the service is provided by the resource. Thus, an array may be thought of as setting a goal for work item processing and resource treatment, and then matching work items to resources in a way that optimizes the workflow to bring the maximum value to all stakeholders in the workflow. Can be. For example, the arrangement represents the weight for each skill required by the work item, and the business value is calculated by summing the product of the skill level of the resource in each skill and the weight for that skill of the work item. Also by way of example, the arrangement may be a process that is being given for resources and work items, for example, for resources, the time since the resource was available, the time the resource did not use to service the work item, and the work. Measuring the effect of providing services to items on the purpose of the resource; It also tracks the time for which the work item is waiting for service, the evaluation time for which the work item should wait for service, and the time for which the work item has exceeded the target waiting time for the work item. Values for resources and work items are calculated from these processes. For example, the array represents a given weight for each treatment, where the value for the resource is calculated by summing the product of each treatment for the resource and the weight of that treatment, and the value for the work item is calculated for each treatment of the work item and its It is calculated by summing the product of the weights of the treatments.

Preferably, the calculated values are all scaled, i.e. standardized, to ensure that the degree of effect on the final combined value remains within a predetermined boundary. Therefore, determining the business value is preferably a weighting that is a business value weighted by a business value weight that is common to all available or determined resources under resource surplus conditions or common to all available or determined work items under work item surplus conditions. Involves determining business value. Determining the value for the resource also includes determining a weighting value for the resource, which is preferably the value for the resource weighted by a resource value weight common to all available or determined resources, Determining the value includes determining the weighting value for the work item that is the value for the work item weighted by the work item value weight common to all available or determined work items. Selecting a resource or work item includes selecting a resource or work item that has the best combination of weighted value and weighted business value (eg, a summation) for the resource or work item.

Depending on the implementation, the present invention provides one or more of the following advantages.

Allow a virtually unlimited number of skills (limited only by computational power)

· Provides individual weighting of multiple technologies required to process each work item

Consider both required and unneeded skills

Weighting by considering multiple skill levels across multiple technologies

· Each work item can receive different priority weights for business value, resource fairness, and work item fairness

· Each work item can receive different priority weights for resource fairness in different categories

· Each work item can receive different priority weights for work item fairness in different categories

· Allows extensive customization of the processing of matching resources to work items, even on the basis of specific work items or specific resources

The evaluated "service delivery time" is calculated for technical items that require multiple skills in the service based on the weighting of technical requirements, the number of work items waiting for service, and the technical requirements, and the " Serving time "is already distributed to resources

· “Service Delivery Time” calculation based on a single vector, allowing easy distribution of workflows across multiple sites and concentration of site selection decisions.

Although the invention has been characterized in one way, it also includes an apparatus for carrying out the method. The apparatus preferably comprises an effector for each method step—in contrast to the means, an entity that performs the corresponding step. The invention also includes a computer-readable medium containing instructions for causing a computer to execute the method steps when executed on a computer.

These and other features and advantages of the present invention will become more apparent from the description of exemplary embodiments of the invention considered in conjunction with the drawings.

1 is a block diagram of a job processing system incorporating embodiments of the present invention.

2 is a block diagram of work-item classification in the system of FIG.

3 is a block diagram of resource entitlements in the system of FIG. 1;

4 is a functional flow diagram of the operation of the selection engine in the system of FIG. 1 upon surplus of resources.

5 is a functional flow diagram of the operation of the selection engine in the system of FIG. 1 upon redundant work items.

* Explanation of symbols for main parts of the drawings

100: work item 102: registration maker

104: resource 106: selection array

110: Work Item Sorter 112: Waiting Work Item List

114: Selection Engine 116: Resource List

118: resource status update 120: resource list of idle status

1. Environment

1 shows a work item 100 waiting to be processed, a resource 104 for processing the work item 100, and assigning each work item 100 to each resource 104 for processing. A match-maker 102, and a selector 106 that determines which work item 100 should be assigned to which resource 104 and controls the match maker 102. do. By way of example, the processing system can be a call center, where work item 100 has a standby communication, resource 104 has an agent and an interaction response system, The match maker 102 has an automatic communications distributor (ACD), and the selection arrangement 106 also has a control program located in the processor accompanying the ACD or in the memory of the ACD and executed on the processor of the ACD. do. As described, the job processing system of FIG. 1 is conventional.

2. Selection Array Algorithm

The purpose of the selection arrangement 106 is to mate the work item 100 and the resource 104 in such a way as to bring the maximum value to all stakeholders in the work processing system. This includes customers, resources, managers, and so on. In accordance with the present invention, to perform this optimization, the selection arrangement 106 evaluates the work item 100-to-resource 104 matching of the three components.

2.1 Three System Values

2.1.1 Business Value (BV)

The business value (BV) of work item 100-to-resource 104 matching is determined for work item 100 that is addressed based on the resource 104 skills and technical proficiency and the technical requirements of work item 100. Resource 104 is a measure of qualification.

2.1.2 Resource Treatment Value (RTV)

The resource processing value (RTV) of work item 100-to-resource 104 matching is compared with each resource 104 purpose as well as a measure of how the resource 104 spends time compared to other resources 104. to be.

2.1.3 Work Item Treatment Value (WTV)

The work item processing value (WTV) of the work item 100-to-resource 104 matching is compared with other work item 100 as well as each processing purpose for a particular work item 100 to compare work item 100. This is a measure of how it is processed.

2.2 Two Basic System States

The evaluation of the three system values described above depends on the state of the work processing system in the evaluation. The following describes the effects of the two possible states and the application of three system values on the system.

2.2.1 Resource Surplus (RS)

Idle resources waiting for work items because there are no work items 100 waiting or no work items 100 waiting for resources 104 in idle state to have a "required" skill. If 104, the system is considered to be in a state of resource surplus (RS). In the case of resource surplus, when work item 100 arrives at the system, the goal is to find the best resource 104 to deal with work item 100. Thus, the resource processing value and business value of each resource 104 that matches the new work item 100 is evaluated to find a resource 104-to-work item 100 match that brings the best value to the organization. In the case of resource surplus, work item processing is not a problem since work item 100 receives immediate processing if resources 104 with the required technology are available.

2.2.2 Work Item Surplus (WS)

There is no idle resource 104 in the system, or there is no idle resource 104 with the technology required to service the waiting work item 100 and the work item 100 waiting for the resource 104 assignment. ), The system is considered to be in the state of work item surplus (WS). In the case of work item surplus, when the resource 104 becomes available to handle the work item 100, the goal is to find the best work item 00 that the resource 104 will handle. Thus, the resource 104-to-work item 100 that evaluates the work item processing value and the business value of each work item 100 that matches the newly available resource 104 to bring the best value to the organization. Find the match. In the case of work item surplus, if there is an available work item 100 with the skills required by the resource 104, the resource handling method is problematic because the work item 100 is received immediately when the resource 104 becomes available. Does not become.

2.3 Calculation of Business Value

_The matrix A _{availresources, maxskills,} where _{availresources} is the number of resources 104 available and _maxskills is the maximum number of skills defined in the system _{, shows} that A _{n, m} is the skill level for resource n and skill m (e.g., 0 and 10 (Integer value in between) is assumed to represent available resource 104 and associated skill level.

Also, _{pendingworkitems} the number of work items (100) that is not provided by the system services and _maxskills the maximum number of the techniques defined in the system, the matrix BR _{pendingworkitems, maxskills} is BR _{n, m} this technique weights for work items n and technical m Assume that (an integer between 0 and 10) is represented to represent the description weight of the work item 100.

Finally, _{pendingworkitems} a number of system services that are not available work items (100) and _maxskills the maximum number of the techniques defined in the system, the matrix BRR _{pendingworkitems, maxskills} the resources 104 in the technology m to the handle a work item n Assume that BRR _{n, m} (true, false) indicates whether the skill level> 0 should be indicated to indicate the requirements of the description for the work item 100. This array is not used in the calculation of business value, but more later.

2.3.1 Resource Surplus

In the case of a resource surplus system state, the business value of the resource 104 is calculated as follows:

2.3.2 Work Item Surplus

For work item surplus system status, the business value of work item 100 is calculated as follows:

2.4 Resource Resource Value Calculation

_The matrix T _{availresources, 3,} the number of resources 104 waiting for _{availresources} to service work item 100 in the system, is T _{n, 1} = time in seconds since it became available, T _{n, 2} = work item ( The percentage of login time not consumed to service 100 (work time of work item and work after associated call of work item), and T _{n, 3} = service of work item 100 available for this resource. Assume that you measure the activity of each resource during the current login session by measuring how much you move it (see the section on calculating and setting resource purpose). If desired, other and / or further treatment measures may be used.

Next, TW _{pendingworkitems, 3} is the weight applied to the process value T of resource 104 such that TW _{n, m} is a weight for resource n applied to T _{n, m} ∀m (where 1 ≦ _m ≦ 3). Assume that

Now, the resource processing value of the resource 104 is calculated as follows:

2.5 Calculation and setting of resource objectives

To satisfy the following placed into a vector containing the G _maxskills, percentage allocation for the purpose _numresources each resource 104 to _maxskills.

Now, TT _{numresources, maxskills} is placed into a vector containing the total time spent by each resource 104 in the processing of work item 100. This vector is updated as follows:

1. The resource r ends processing of the work item wi.

2. The work item wi has a description weight vector BRwi.

3. The total processing time Tp is recorded.

4. The TT is updated as follows:

TT _{r, n} = TT _{r, n} + T _p , ∀n, where 0 <n <maxskills and BR _{wi, n} > 0.

Now, the total time spent processing each technique is available for each resource 104. This total time, along with the expected new work item 100 and percent allocation purpose, calculates whether processing the work item 100 would benefit the resource 104 when moved toward the allocation purpose as follows: Can be used to:

Set PA _{n, maxskills} to the actual percentage allocation for resource n to satisfy:

Now, the current distance of resource n from the purpose is calculated as follows:

Now set T _t to the total time that resource 104 has processed work item 100 and T _wi to the total number of work items 100 processed by resource 104. At this time, T _avg = T _t / T _wi . Finally, let TTE _{n, maxskill be} the total time evaluated and let PAE _{n, i} be the percent allocation evaluated for resource n when processing work item wi _new with technical vector BR _new :

TTE _{n, i} = TT _{n, i} + T _avg , ∀i, where 0 <i <maxskills and BR _{new, i} 〉 0

Now, the estimated distance of resource n from the objective is calculated as follows:

Now, the degree of improvement versus the objective can be expressed as follows:

Improvement = DG _n -DGE _n

Here, a negative value indicates moving away from the objective, and a positive value indicates moving in the target direction. However, this results in some ambiguous calculations in later calculations. Thus, the improvement measure is shifted to ensure a positive value as follows:

Improvement = T _{n, 3} = DG _n -DGE _n + 2 x maxskills

This yields a value between 0 and 4 x maxskills. This value is adjusted in later calculations for practical use.

2.6 Calculating Work Item Processing Values

Matrix C _{pendingworkitems, 3,} which is the number of work items (100) that pendingworkitems is waiting for service on the system, where C _{n, 1} = time in seconds that work item (100) waits for service, C _{n, 2} = work item (100) The evaluation time in seconds to wait for this service (see the section on calculating evaluated service waits), and C _{n, 3} = the time in seconds (or target wait time) that the work item 100 is waiting for targets. If not, it is assumed that 0) represents the current waiting condition of each work item. If desired, additional and / or other treatment measurements may be used.

In addition, it is _{assumed that} CW _{pendingworkitems, 3} represents the weight applied to the process value C by making CW _{n, m} a weight for the resource n applied to C _{n, m} ∀m (where 1 ≦ _m ≦ 3).

Now, the resource disposal value of the work item 100 can be calculated as follows:

2.7 Calculation of evaluated service waits

The calculation of the estimated service waits consists of a moving average of past wait times, weighting across the technology categories, and technical requirements as follows:

First, set F _maxskills as a vector holding a moving average of past wait times (one moving average for each technology up to _maxskills ), AW as the average number of work items 100 waiting for service, and also waiting for TWWI The total number of work items (100). Now, whenever a work item wi having a wait time WT is assigned to resource 104, F and AW are updated by:

So, this is not only the moving average, but also the variable moving average of the new work item 100 that contributes to the moving average between 0 and 10 percent based on skill needs and weights.

Now, calculating the evaluated service wait EWS based on the given work item wi and the technical requirement vector BRR is done as follows:

Under normal operating conditions, the TWWI / AW is close to one. However, if the workload spikes or drops suddenly, it will adjust the assessment based on the slow moving average to more accurately reflect the state of the system.

3. Decision

The final task is in the selection of resource 104 or work item 100. In the case of resource surplus, it is the weight of business value versus resource processing value, and in the case of work item surplus, it is the weight of business value versus work item processing value. W _RSBV = weight of business value in resource surplus, W _AT = resource processing value in resource surplus, W _WSBV = weight of business value in work item surplus, and W _CV = weight of work item processing value in work item surplus. . These are manageable values.

3.1 Resource Surplus and Resource Selection

In the case of resource surplus, an attempt is made to maximize the business value and resource processing value according to the specified weights by selecting the resource 104 with the highest score as follows:

Max [(RSBV _n x W _RSBV ) + (RTV _n x W _AT )}, where A _{n, i} 〉 0 ∀BRR _{l, i} 〉 0

3.2 Work Item Surplus and Work Item Selection

In the case of work item surplus, an attempt is made to maximize the business value and work item processing value according to the specified weights by selecting the work item 100 with the highest score as follows:

Max [(WSBV _n x W _WSBV ) + (WTV _n x W _CV )}, where A _{l, i} 〉 0 ∀BRR _{n, i} 〉 0

3.3 Options: Dynamic Business Value and Work Item Handling Weights

In the case of work item surplus, the option of setting a certain weight for the business value and work item recipe value is to allow the value to change over time. This change increases the weight of the work item processing value over time to prevent the needy state and move closer to the target service time of the work item. In this scenario, work item 100 has initial values of W _RSBV = 1 and W _CV = 0. These values may change over time (during each assessment) as follows:

W _cv = (current wait time) / (target wait time)

If either is greater than 1,

W _RSBV = 1-W _CV .

This starts the calculation with the maximum weight for the description of the resource 104 and is moved to apply the maximum weight to ensuring a fair treatment of the work item 100.

4. Scaling the Results for Real Application

The algorithm selects resource 104 or work item 100 based on the highest score. However, it may be impractical to determine the weight of each value that would provide an optimal solution for the system. For example, if resource 104 is scored very high in work value (10 x 10 = 100), but work item 100 waits for an average of five minutes, work item processing value (WTV) is around 300. Can be tall. So, with work item 100 age, WTV has an increased effect on overall choice. This is not desirable.

So, to be substantially more applicable, the calculation of each selection value must be scaled. Scaling values are needed in each calculation executed for all resources 104 or all work items 100 as follows (scaling values are represented by S):

Now, the scaled values for RSBV, WSBV, RTV, and WTV can be expressed as follows:

The final result of the calculation is that each of the scaled values WRSBV, WWSBV, WRTV, and WWTV are between 0 and 100. Now, regardless of the scaling interval or unit of measurement for the initial configuration of each value, the weighting of these values in the overall resource 104 and work item 100 selections can be consistently expressed.

Note that the weighting of the components of the resource processing value T _{n, i} and the work item processing value C _{n, i} is also scaled by the calculation.

5. Examples

An example of resource surplus (resource selection) work item 100-to-resource 104 matching is shown to illustrate the concepts defined above. Suppose we have two resources 104, Jane and Jeff. It also has skills 1-6, that is, English, Chinese, technical support, sales, PC, and UNIX. Jeff is fluent in English (technical level 10), speaks Chinese (technical level 5), and also trained in technical support (technical level 8) and UNIX (technical level 10). Jane speaks English (technical level 10), is trained and experienced in sales (technical level 10), is just learning technical support (technical level 2), and also has a PC (technical level 7) and UNIX (technical level). 6) have experience. Also assume that both Jeff and Jane are logged in and waiting to receive work item 100.

Now, matrix A of available resources is as follows:

Jeff also assumes he has been idle for 1 minute and is 75% idle that day. Jane has been idle for three minutes and is 50% idle that day. Now, the matrix T is as follows:

Now, work item 100-call-arrives at the system. Work items have the following requirements: English (weight 10), technical support (weight 10), and UNIX (weight 5). All three of these techniques are required. Now, the matrix BR and matrix BRR are as follows:

BR = [10 0 10 0 0 5]

BRR = [1 0 1 0 0 1]

Finally, when the work item 100 arrives, it is assigned the following weights:

Now, we can calculate the best choice as follows:

Total value for Jeff is:

100 (.2) + 22.92 (.8) = 40.33

The total value for Jane is:

65.22 (.2) + 100 (.8) = 93.044

Therefore, Jane is chosen to handle the work item. Note that Jeff has a higher score in business value and Jane has a higher score in resource equity value. However, due to the 80% weighting of resource equity values, work items were given to Jane.

6. Realize optional arrangement

6.1 Resource Items

In order to evaluate the resource 104 and assign it to the best work item 100, it must be given to the system according to a particular service function. Therefore, the selection arrangement 106 of FIG. 1 includes a list 116 of all resources 104 and their entitlements in the system of FIG. These qualifications include relatively static data as well as dynamic data and are shown in FIG. 2. The entitlement 200 for each resource 104 includes a description vector 202 that is a list of all the skills owned by the resource 104. The entitlement 200 for each resource 104 is also defined by the technology-level vector 204, which indicates the level of this resource 104 in each technology owned by this resource 104, and this resource 104. It includes a skill assignment purpose vector 206 that represents a given assignment purpose for all owned skills. These are relatively static management values. The entitlement 200 also includes a descriptive time vector 208 that represents, for each skill, the percentage or total amount of time spent by this resource 104 when processing work items 100 that require this skill. It includes the same measure (if the work item 100 processed by the resource 104 requires multiple descriptions of the resource 104, the description time of each required technology is determined by the resource 104 being determined by the work item). 100). The measurement also includes a work time vector 210 representing one or more of the following: the total time the resource 104 was logged in, the resource 104 was associated with the customer. Total time spent in live communication (in-call time), total time other than call waiting time spent by resource 104 to process a call (ACW time), resource 104 ) The total time busy, the total time resources 104 prepared to process work item 100, Source 104 of time that are consumed in processing the work item (100) (TP). The measurements of the working time vector 210 correspond to similar measurements conventionally maintained for agents in the ACD system and are calculated in the same manner. The entitlement 200 also includes state information 212 that indicates the current state of the corresponding resource 104 and the last state change time of the resource 104. The latter value is used to determine the amount of time that the resource 104 has been idle since processing last the work item 100. The resources 104 in idle state form an idle resource list 120, which is a subset of the resource list 116.

6.1.1 unified technology

Unity skills are individually independent skills. English, Spanish, Sales, Technical Support, UNIX, Windows, and Macintosh fall into this category. These are also the simplest for the manager. Each description is specified by a number starting at 1 such that A _{n, m} represents the description level of resource n for description m.

6.1.2 Collective Description

Collective techniques are a bit more difficult to work with. These are techniques that consist of a combination of the one-way techniques described above. Collective techniques are necessary in that there are resources 104 that can handle several techniques individually or only in specific combinations. For example, suppose a call center has an agent who speaks Spanish and knows UNIX, Windows, and Macintosh. It is also assumed that this agent has also undergone some technical support training and is therefore skilled in technical support. However, this agent is getting used to the role of technical support. During the first two weeks, the agent should only handle Spanish technical support calls to Windows. During the next two weeks, the agent can also handle Spanish technical support calls for Macintosh. Finally, the agent is also allowed to handle Spanish technical support calls for UNIX. This scenario is most easily managed with three collective technologies: Spanish Windows Technical Support, Spanish Macintosh Technical Support, and Spanish UNIX Technical Support.

Now, more generally, the collective description is managed as follows. Assume some collective description g is a collection of unitary descriptions a, b, and c. Now, A _{n, g} represents the skill level of resource n for the collective skill g, where g> maxunitskills and maxunitskills is the total number of one-way skills.

This allows the resource 104 to be further differentiated by the ability to handle each technical requirement as well as a combination of technical requirements.

6.2 Update Resource Status

The action items 206-212 of the entitlement 200 of all resources 104 in the resource list 116 are maintained or dynamically updated by the resource status update 118 of FIG. 1. Resource status update 118 provides an interface through which applications can communicate information to update the status of resources 104 as needed. Resource status update 118 receives a notification of a resource 104 status change event and updates the lists 116 and 120 accordingly.

6.3 Work Item Sorter

In order to evaluate the work item 100 and distribute it to the best resource 104, it must be globally classified and given to the system according to specific service needs. This classification is performed by the work item classifier 110 of FIG. The classification 330 needed by the classifier 110 to provide for each work item 100 is described later and shown in FIG. 3.

6.3.1 Eligibility for Calculation of Business Value

The classification 300 of each work item includes a description vector 302 which is a list of all possible descriptions 1-M that the work item 100 may have. Each technique required or simply desired for the processing of work item 100 should then be added to work item classification 300 for each category.

6.3.1.1 one-way description

Each classification 300 includes a BR vector 306 that is a list of skill weights assigned to each skill in the description vector 302 for this work item 100. For each one-way skill associated with work item n, the weight (for example, an integer between 0 and 10) is BR _{n, m} = w (where 1 ≦ _m ≦ maxunitskills, 0 ≦ w ≦ 10, and maxunitskills is the Number).

6.3.1.2 collective description

BR vector 306 includes both one-way and collective descriptions. For each collective technique associated with work item n, the weight (an integer between 0 and 10) is BR _{n, m} = w where maxskills≤m≤maxskills, 0≤w≤10, and maxskills is a combined collective technique And the total number of unitary descriptions).

6.3.1.3 Technical requirements

Each classification 300 includes a BRR vector 304 indicating which of the descriptions in the description vector 302 are required by this work item 100. For each collective and one-way description associated with work item n, a true or false value must be specified in BRR so that BRR _{n, m} indicates whether description m is required for work time n. Vectors 302, 304, and 306 together represent the business value of work item 100. Note that it is possible that one technology contributes to the business value calculation but is not the required technology. For example, suppose some agents in a call center are trained in valid communication, but this is not required to handle the call. These agents are trained and therefore score high in business value. However, if the technique does not have a true value in the requirement vector BRR but a weight greater than zero in the BRR, then the untrained agent is not excluded from the selection process.

6.3.2 Qualification Weights for Resource Equity Values

Each classification 300 includes a qualification weight (QWRFV) vector 308 for resource fairness values. The value of TW _{n, m} for resource n should be filled with one of the following three for fairness value (1≤m≤3): TW _{n, 1} is the weight given to time since resource n becomes available (Or is it the idle time?), TW _{n, 2} is a weight given to the percentage of time that resource n did not spend processing work item 100, and TW _{n, 3} is the current work item 100's The weight given to how much processing has moved the resource 104 towards the service object. Although the value of TW can be any value, the percentage weight of each value for some resource n It is most appropriate to think of it as. In the case of a simple default, work item 100 is expected to have only one set of TW values in common for all resources 104.

6.3.3 Qualification Weights for Work Item Fairness Values

Each classification 300 includes a qualification weight (QWWIFV) vector 312 for work item fairness values. As in the above resource, the value of CW _{n, m} for work item n should be filled with one of the following three for fairness value (1≤m≤3): CW _{n, 1} is determined by work item 100 The weight given to the time spent waiting for processing, CW _{n, 2} is the weight given to the estimated total wait time of the work item _, and CW _{n, 3} is the wait time for the work item 100 to exceed the target service time. The weight given. Although the value of CW can be any value, the percentage weight of each value for some resource n It is most appropriate to think of it as.

6.3.4 Qualification Weights for Business Value

When the classifier 110 communicates a new work item 100 to the system for distribution, it is assumed that the classifier 110 does not know the state of the system at all, and therefore should provide enough information to determine in any state.

6.3.4.1 Resource Surplus

Each class 300 provides a resource surplus weight 314 that includes a weight of business value W _{RSBV in} resource surplus and a weight of resource fairness W _AT in resource surplus, so that the resource surplus can be determined.

6.3.4.2 Work Item Surplus

For classification in the work item surplus, each classification 300 provides a work item surplus weight 316 including the weight of the business value in the work item surplus and _WSBV , and the weight of the work item fairness in the work item surplus and _CV . do.

6.3.5 Eligibility for Target Wait Time

Finally, each classification 300 provides a target latency target 318 for the work item 100.

6.3.6 Classification Generation

For example, values 308-318 are pre-managed for other business values (values of fields 302-306), so classifier 110 only needs to determine business values (values of fields 302-306). Only—for example, in the same manner as in the prior art-based ACD system—filling fields 308-318 with the corresponding management values.

6.4 Selection Engine

4 and 5 show an embodiment of the selection engine 114. 4 shows the operation of the engine 114 when the sorted work item 100 becomes available, and FIG. 5 shows the operation when the resource 104 becomes available.

6.4.1 Resource Surplus

In step 400, when the sorted work item 100 is in the standby work item list 112 or becomes available, the engine 114 causes any resource 104 to process the work item in step 402. The resource list 120 in idle state is checked to determine if it is available. If idle resource item 120 is empty, engine 114 returns work item 100 to list 112 at step 410 and operates until resource 104 becomes available at step 412. To exit. If the idle resource list 120 is not empty, then the engine 114 determines the work item classification 300 to determine if the resources 104 available in step 414 have the skills required by the available work item 100. Compares the skill requirement 304) with the skill level 204 of the available resource entitlement 200. Otherwise, engine 114 proceeds to step 410, in which case engine 114 proceeds to step 420.

In step 420, engine 114 calculates a business value (RSBV) for each available resource 104 having the skills required by the available work item 100. The calculation is performed as described in section 2.3.1 above. The engine 114 then calculates a resource processing value (RTV) for each available resource 104 having the skills required in step 422. The calculation is performed as described in sections 2.4 and 2.5 above. The engine 114 then selects the available resources 104 based on the weighted RSBV and RTV scores combined in step 424. This calculation and selection is performed as described in section 3.1 above. The engine 114 then removes the selected resource from the idle resource list 120 at step 426, and designates the work item 100 available to the selected resource 104 by the match producer 102 at step 428. And also terminates operation at step 430 until another classification work item 100 or resource 104 becomes available.

6.4.2 Work Item Surplus

In step 500, when the resource 104 is in the idle resource list 120 or becomes available, the engine 114 determines whether any work item 100 is available for processing in step 502. Check the pending work item list 112 to verify. If the queued work item list 112 is empty, the engine 114 returns the resource 104 to the list 120 at step 510 and until the work item 100 is available at step 512. End the operation. If the queued work item list 112 is not empty, then the engine 114 determines the available work item classification 300 to determine if the description of the resource matches the technical requirements of the available work item 100 in step 514. Compare the skill requirements 304 of the skill level 204 of the resource entitlement 200. Otherwise, engine 114 proceeds to step 510, in which case engine 114 proceeds to step 520.

In step 520, engine 114 calculates a business value WSBV for each available work item 100 for which the technology needs to match the technology of resource 104. The calculation is performed as described in section 2.3.1 above. The engine 114 then calculates a work item recipe value (WTV) for each available work item 100 for which a description needs to match the description of the resource 104 in step 522. The calculation is performed as described in sections 2.6 and 2.7 above. The engine 114 then selects the available work item 100 based on the weighted WSBV and WTV scores combined in step 524. This calculation and selection is performed as described in section 3.2 above. The engine 114 then removes the selected work item 100 from the pending work item list 112 in step 526 and assigns the resources 104 available to the selected work item 100 in step 528. It requests the match maker 102 and ends the operation until another resource 104 or sorting work item 100 becomes available in step 530.

6.5 variant

Of course, various changes and modifications will be apparent to those skilled in the art in the above-described embodiment. For example, the concise nature of the algorithm allows the distribution of the system over multiple sites. If updating the current value of the expected vector F is very effective for many sites, the average number of waiting work items, AW, the number of work items currently waiting, TWWI, and a vector representing the serviceable technology (can provide services). To prevent delivery of work items to missing sites). This central location has enough information to distribute work items across multiple sites based on estimated service times. Alternatively, this information may be initially delivered to the appropriate location by an external process that accesses this central location (ie, a pre-route solution). Such changes and modifications may be made without departing from the spirit and scope of the present invention and without obscuring the accompanying advantages. Therefore, such changes and modifications are intended to be covered by the following claims, except as limited by the prior art.

The present invention provides a new workflow resource-selection and work item-selection arrangement that attempts to match work items with resources in a manner that brings maximum values to all stakeholders in the workflow.

Claims (10)

A method comprising selecting a resource for a work item or selecting a work item for a resource, the method comprising: The step of selecting a resource for a work item is Determining available resources that possess skills required by the work item; For each of the determined resources, determining a business value that causes the resource to service a work item; For each of the determined resources, determining a value for a resource providing a service to a work item; And Selecting the determined resource having the best combination of business value and value for the resource, to provide service to the work item, Selecting a work item for a resource Determining available work items requiring skills owned by the resource; For each of the determined work items, determining a business value that causes a resource to service the work item; For each of the determined work items, determining a value for the work item to be serviced by the resource; And Selecting the determined work item having the best combination of business value and value for the work item, such that the service is provided by the resource. The method of claim 1, wherein determining the business value that causes the resource to service the work item comprises: A business weighted by multiplying (a) the business value weight corresponding to the work item and (b) the product of the level of each required skill of the resource and the product of the weight of the required skill of the work item. Determining a value; Determining the value of a resource (c) determining the weighted resource processing value value by multiplying the sum of the product's treatment weight corresponding to the work item and (d) the product of each processing value of the resource and the weight of the processing value of the resource. With; Also The step of selecting the determined resource Selecting the determined resource having the best combination of weighted business value and weighted resource processing value value. The method of claim 1, wherein determining the business value that causes the resource to service the work item comprises: For each of the resources determined, comprising a sum of the skill level of the resource in the skill and the skill weight of the work item for the skill across all skills to determine business value; Also Determining the value of a resource For each of the determined resources, determining the resource processing value by providing a sum of the product of the resource value for the resource processing value and the weight of the work item for the resource processing value over all the resource processing values; Characterized in that the method. The method of claim 1, wherein determining the business value that causes the resource to service the work item comprises: The weighted business value is determined by multiplying (a) the business value weight corresponding to the work item and (b) the product of the level of each required skill of the resource and the weight of the weight of the required skill of the work item. Providing a step; Determining the value for a work item determining the weighted work item processing value value by multiplying the sum of the product of the work item processing value weights corresponding to the work item and (d) the weight of each processing value of the work item and the processing value of the work item. A step; Also The steps to choose Selecting the determined work item having the best combination of the weighted business value and the weighted work item processing value value. The method of claim 1, wherein determining the business value that causes the resource to service the work item comprises: For each of the determined work items, comprising the sum of the skill level of the resource in the skill and the product weight of the work item for the skill over all skills to determine the business value; Also Determining the value for a work item For each determined work item, the sum of the product of the work item value for the work item processing value and the weight of the work item for the work item processing value is provided over all work item processing values to determine the work item processing value value The method comprising the step of. The method according to claim 3 or 5, wherein the resource processing values of the resource include the time since the resource became available, the time spent by the resource not providing service to the work items, and the service item on the work item. To have a measure of the effect it has on the purpose of the resource, The work item processing values of the work item are characterized by having the time spent waiting for the work item to be serviced, the time spent waiting for the item to be serviced, and the time the work item has exceeded the target waiting time. How to. 7. The method of claim 6, wherein the measure of effectiveness comprises: (a) the distance of the actual allocation of the work time of a resource from among the techniques from the purpose allocation of the work time of the resource among the techniques if the resource serves the work item from the purpose allocation; (b) including the difference between the distances of the estimated allocations of working time of the resources among the techniques, The estimated waiting time for a work item to wait for service is (a) the ratio of the total number of work items waiting for service to the average number of work items waiting for service, and (b) the weight of each technology and the skills required. And the product of the sum of the mean wait times of each of the required techniques, each weighted by a ratio of the sum of the weights. 6. The method of claim 3 or 5, wherein determining the business value for each of the determined resources Determining a scaled business value with a scaled business value by a first scaling factor common to all determined resources; Determining the value of the resource processing value for each of the determined resources For each resource processing value, determining the scaled value of the resource with the value of the resource for that resource processing value scaled by a scaling factor common to all resources determined at that resource processing value, and Scaled sum by a second scaling factor common to all determined resources, over all resource processing values of the product of the scaled value of the resource for the resource processing value and the weight of the work item for the work processing value. Determining the scaled resource processing value value; The step of selecting the determined resource Selecting the determined resource having the best sum of the scaled business value and the scaled resource processing value value to provide service to the work item; or Determining business value for each of the determined work items Determining a scaled business value with a scaled business value by a first scaling factor common to all determined work items; Determining the work item processing value value for each determined work item For each work item processing value, the scaled value of the work item with the value of the work item for that work item processing value scaled by a scaling factor common to all work items determined in that work item processing value. Determining, and By a second scaling factor common to all work items determined, over all work item processing values of the product of the scaled value of the work item to work item processing value and the weight of the work item to work item processing value Determining a scaled work item processing value value with a scaled sum; The step of selecting the determined work item is Selecting the determined work item having the best sum of the scaled business value and the scaled work item processing value value to be serviced by the resource. An apparatus comprising the method of any one of the preceding claims. A computer-readable medium, when executed in a computer, comprising instructions for causing the computer to execute the method of any one of claims 1 to 8.